From the Board of Governors: AI Presents Opportunities and Challenges for Medical Specialty Boards
By ABR President-elect John A. Kaufman, MD, MS; ABR President Cheri L. Canon, MD; and ABR Executive Director Brent Wagner, MD, MBA
August 2026;19(4):3
Discussions regarding the accelerating introduction of artificial intelligence (AI) are ubiquitous in the context of healthcare delivery, including in the specialties certified by the ABR. The expanding breadth of potential applications, and their far-reaching and unpredictable consequences, challenges organizations and institutions that are tasked with preparing for a landscape that is rapidly shifting.
Unsurprisingly, ABR board members and senior staff are often asked about the organization’s view of what AI might mean to our business functions and the certification process.
Focusing on the certification processes, we engaged an IT consulting firm to provide an outside perspective on the challenges and opportunities that AI presents for medical specialty certification and related credentialing activities. One of the initial phases of the engagement involved interviews with several volunteer board members and staff. Additionally, we benefited from the participation and perspectives of three diplomates — Dr. Florence Doo of the University of Maryland, Dr. Christoph Wald of Mayo Clinic Rochester, and Dr. Grace G. Zhu of the University of Utah — whose viewpoints and experience provided invaluable objectivity and a broad scope on the diverse issues under discussion.
Although our Board is always looking at ways to gain operational efficiencies and reduce costs, we specifically excluded standard “business functions” from this engagement. We also considered paradigm shifts in clinical practice to be out of scope, including, for example, the implications of independent interpretation algorithms and associated agency supplanting the work of a diagnostic radiologist. Our emphasis, instead, was on internal guidelines for ABR staff and volunteers for ethical and defensible use of the capabilities afforded by AI as applied to certification activities. Foundational limits would involve, for example, an inviolable requirement to have score determinations and certification decisions within the sphere of human responsibility. We are also prohibiting the use of generative AI to produce exam content, including images.
Within the framework of some fundamental constraints, the Board remains open to considering enhancements that may make the programs and processes that support certification more efficient and more valid based on exam paradigms that have yet to be developed. As these innovations mature, the ABR’s position is to maintain a high level of transparency to continue to preserve our legitimacy as a certifying body. Further, pilot programs, subject to human oversight, would be conducted in low-risk areas that could improve blueprint coverage of the domain or the quality and relevance of the pool of exam content.
Medical specialty boards are generally expected to “stay the course” and act conservatively when making decisions regarding change or innovation. This is especially challenging when faced with the fast pace of technological change that simultaneously opens new opportunities and untested frontiers of risk and downsides. Our goal is to balance those factors, and to adapt accordingly while maintaining certification programs that are both credible and valuable to our candidates, diplomates, and the public. For more information, please see A Few FAQs About the ABR and AI.
